1. Field of the Invention
The invention relates to a solid state image pickup device and a camera and, more particularly, is suitable for use in a CMOS area sensor.
2. Related Background Art
In recent years, the CMOS area sensor in which a photodiode and a MOS transistor are formed as one chip is used as a solid state image pickup device. The CMOS area sensor has such advantages that electric power consumption is smaller, a driving electric power is smaller, and a higher processing speed can be realized as compared with those of a CCD. The general CMOS area sensor is constructed by forming a plurality of pixels in a matrix shape, in which each pixel has: a photodiode; a floating diffusion (hereinbelow, also abbreviated to FD as necessary) region; a transfer transistor for transferring carriers from the photodiode to the FD region; and a reset transistor for resetting the FD region to a predetermined electric potential.
A technique regarding the CMOS area sensor whose dynamic range is widened has been examined (for example, refer to Shigetoshi Sugawa, and other five persons, “A 100 db Dynamic Range CMOS Image Sensor Using a Lateral Overflow Integration Capacitor”, ISSCC 2005/SESSION19/IMAGES/19.4, DIGEST OF TECHNICAL PAPERS, 2005 IEEE International Solid-State Circuit Conference, Feb. 8, 2005, p 352-353, 603). According to the CMOS area sensor in the above non-Patent Document, in each pixel, further, a capacitor region whose capacitance is larger than that of the FD is formed, one terminal of the capacitor region is connected to the FD through a switch, and the other terminal of the capacitor region is connected to the ground. Thus, when carriers overflow from the photodiode by strong light, the overflowed carriers are held into the capacitor region, thereby enabling a signal corresponding to a quantity of overflowed carriers to be outputted and widening the dynamic range.
In the CMOS area sensor, however, there is a problem that a color mixture with the adjacent pixel occurs irrespective of the presence or absence of the widening function of the dynamic range mentioned above. FIGS. 8A and 8B are diagrams showing a mechanism of the occurrence of the color mixture according to conventional pixel layouts. In FIG. 8A, a transfer unit 502 and a MOS unit 503 are arranged under a photodiode 501. When the pixels having such a layout are arranged, even if a device separating region is provided between the pixels, there is a case where the color mixture cannot be avoided. In the case of FIG. 8A, since the MOS unit 503 is arranged in the vertical direction or the like, a distance between the photodiodes of the adjacent pixels is larger than that of the adjacent pixels in the lateral direction. However, for example, as shown in FIG. 8A, photodiodes 501 and 504 are arranged in the lateral direction so as to sandwich only the device separating region. There is, consequently, such a problem that the color mixture occurs because the carriers which have been photoelectrically converted in a deep layer portion of silicon leak or light or the like which has obliquely entered and has been reflected by an aluminum layer or the like enters the adjacent photodiode.
As shown in FIG. 8B, a transfer unit 512 and a floating diffusion region 513 are provided on the right side of a photodiode 511 and the MOS unit is provided under the photodiode 511. Owing to such a layout, although a distance between the photodiode 511 and a photodiode 521 of the pixel which is adjacent on the right side of the photodiode 511 is larger than that in the case of FIG. 8A, since a distance between the floating diffusion region 513 and the photodiode 521 is small, there is a problem that the carriers leak into the floating diffusion region 513 and the color mixture occurs.
According to the layout disclosed in the above non-Patent Document, since it is necessary to set a large capacitor region, it exercises a large influence on the problem of the color mixture in dependence on the layout.